From Consensus to Individual Differences: Typicality Links Brain and Behavior Across Naturalistic Contexts
Year: 2026
Type: preprint
Abstract: When people experience the same naturalistic events, their brains and behavior converge on remarkably similar responses. This shared response is widely used to identify what people have in common, but whether the consensus itself carries meaning has remained largely unexplored. Here we propose that the group consensus is not merely an average, but an approximation of a shared solution to the computation being performed, such that an individual’s proximity to it, their typicality, indexes how well they perform that computation. Across behavioral tasks spanning social and non-social cognition, typicality was a stable characteristic of the individual, yet specific to the computation engaged rather than a general tendency to conform. Neural typicality during movie viewing showed the same organization, and was linked to behavior most strongly where the relevant computation was engaged: typicality in face-selective cortex was more strongly related to behavior on tasks involving faces than on matched tasks without faces, particularly during movies containing faces. Critically, where ground truth was available, proximity to the consensus predicted objective performance: neural typicality measured during passive movie viewing predicted accuracy on independent motion-prediction and face-recognition tasks. These findings suggest that, when individuals share objectives and constraints, the group consensus can serve as an empirical benchmark of successful processing, offering a principled way to measure individual differences in complex behaviors for which no correct answer can be defined. Significance Statement Many complex human behaviors, such as interpreting another person’s emotions, have no single “correct” response, making them difficult to quantify using traditional measures. We propose that the response people converge on when experiencing the same events is not merely an average, but an approximation of a shared solution, so that how closely an individual approaches it, their typicality, can serve as a measure of performance. We show that typicality is a stable characteristic of the individual, specific to the type of information being processed, and linked to the brain systems supporting it. Where correct answers exist, people whose brain responses during movie viewing were more typical performed better on independent tasks. These findings provide a new way to study individual differences in complex, real-world behavior and suggest that shared responses may, under appropriate conditions, reveal how effectively information is being processed.
Institution: Weizmann Institute of Science
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
Award: G8393888960
OpenAlex ID: W7203762758